It has been widely acknowledged that abstraction is central to systematic software engineering. Abstraction is already widely used in software development, from deductive reasoning and formal modeling to the inductive reasoning found in modern data science. We propose that the demands of future software systems require an engineering approach to the systematic creation, maintenance, and use of abstractions; we call this Abstraction Engineering [1]. Importantly, Abstraction Engineering (AE) brings together concepts of abstraction across the inductive/deductive reasoning divide. Our paper outlines the foundations, explores key challenges, presents research questions, and offers a roadmap for future development in this new field.

Our primary goals for this BoF session are:

• Promote the Topic
• Gather Feedback,
• Further Identification of Challenges
• Contribute to AEBoK
• Build a diverse Community.

Anybody who is interested in moving beyond software models to a broader engineering discipline for Abstractions across all types of software-intensive systems, whether deductive and constructed ‘top down’ or inductive and constructed ‘bottom up’ (e.g., through machine learing and AI) are highly welcome to join this BoF session.